SureForge
SkillAn instruction-only skill that makes agents research before they ask, plan before they build, and verify before they call anything done.
- workflow
- verification
- quality
- skill
Anthropic's official GitHub Action that reviews PRs for security vulnerabilities with Claude.
Anthropic's official AI-powered security review GitHub Action. On every PR, Claude analyzes the diff for vulnerability classes — injection, auth and session bugs, secrets handling, unsafe deserialization — and comments with severity-ranked findings and suggested fixes. Complements SAST scanners by reasoning about business-logic risks that rules can't express; instructions and severity thresholds are tunable per repo.
A semantic security pass on every PR — catches what pattern-based SAST misses (broken authz logic, injection via composition) and writes review comments.
GitHub Actions — .github/workflows/security.yml needs `permissions: { pull-requests: write, contents: read }`, then a step `uses: anthropics/claude-code-security-review@main` with `claude-api-key: ${{ secrets.CLAUDE_API_KEY }}`. Local usage: the /security-review command in Claude Code.The README itself warns the action is not hardened against prompt injection — use it only on trusted PRs (e.g. require approval for external contributors). Findings still need human triage, and each PR costs API spend.
Repository
anthropics/claude-code-security-reviewCharted
An instruction-only skill that makes agents research before they ask, plan before they build, and verify before they call anything done.
Guizang's agent skill turns your repo's recent updates into a finished promo video — real components, code-rendered animation, an original code-written soundtrack — no video editor involved.
An agent skill that geolocates a photo with nothing readable in it — OSM geometry, elevation skylines and satellite tiles turn “where was this taken?” into a measured answer with an error radius.